The Kolmogorov-Smirnov test is used to evaluate if histogram distributions deviated from a simulated null hypothesis (H0), comparing to the alternative hypothesis (H1). The distributions with *P* value < 0.05 are normally considered that deviated significantly from a null hypothesis.

Packages required:

- SciPy (installation guide)

*SciPy* (pronounced “Sigh Pie”) is a *Python*-based ecosystem of open-source software for mathematics, science, and engineering.

More detailed usage:

*One sample*

http://docs.scipy.org/doc/scipy-0.15.1/reference/generated/scipy.stats.kstest.html

*Two samples*

http://docs.scipy.org/doc/scipy-0.14.0/reference/generated/scipy.stats.ks_2samp.html

Code:

#!/usr/bin/python
#Performs KS-test
#Usage: python script.py sample1.txt sample2.txt
#Input: distribution values separated by break line in a text file
from scipy import stats
import numpy as np
import sys
x = sys.argv[1]
y = sys.argv[2]
sample1 = np.loadtxt(x,delimiter=",")
sample2 = np.loadtxt(y,delimiter=",")
one_sample = stats.kstest(sample1, 'norm', alternative = 'greater') #null vs alternative hypothesis for sample1. Dont reject equal distribution against alternative hypothesis: greater
two_samples = stats.ks_2samp(sample1, sample2) #sample1 vs sample2
print one_sample,two_samples
#output: sample1 sample2 (D, pvalue) (D, pvalue)

### Like this:

Like Loading...

*Related*